Upgrades

Date Product impacted Description
04 April 2008 L2P AATSR Confidence Flagging Scheme for Medspiration AATSR L2P Files
Medspiration Report No.2 The processing chain for the the AATSR L2P files was significantly updated today :
  • the estimation of the SSES and of each pixel quality flag (proximity_confidence) was completely revised : the former Single Sensor Error Statistics (SSES) scheme applied to AATSR resulted in assigning a quality flag of 2 (bad), 3 or 4 to most of the data, with very few at the highest quality (5). This demonstrated the need for a new scheme adapted to AATSR data, with an improved estimation of the sensor bias and standard deviation  associated with each class of quality. This new scheme was defined by Gary Corlett of University of Leicester and is fully described in this report.
  • a new field has also been added (atsr_dual_nadir_sst_difference) : it contains the difference (in Kelvin) between the estimated dual-view (the SST provided in the sea_surface_temperature field) and the nadir-only sea surface temperature. This information proved to be very significant when estimating the dual-view SST quality.
  • the confidence_flag has also been changed, including some information on the retrieval condition (such as usage of two or three channels)
  • an error in the time value of the pixels (one hour shift) was corrected and all pixels have now a correct time.
04 January 2008 L4 products The processing chain for all L4 products (Mediterranean Sea, North-Western shelves aka Marcoast and Galapagos) was completely revisited. It benefits now on the improvement made within Mersea project on sensor intercalibration and merging, and uses the same processing chain (ODYSSEA) as for the Ifremer global SST analysis. This processing chain is described in the following report.
The format is almost the same as the previous L4 product except that the 'sst_foundation' variable in the netCDF files is replaced with 'analysed_sst' to match the latest GHRSST specification.
The switch between the old and the new ODYSSEA processing chain will occur on 10th of January.
9 November 2007 L4 GAL The area covered by the Galapagos/Coco islands product has been extended to meet the requirements of ESA DIVERSITY project. The area is now 20°South to 20°North in latitude, 120°West to 69°W. 
27 March 2007 All Questionnaire on the use of Medspiration SST data

The Medspiration Service has been operating for nearly two years. It is expected to evolve into a service within the GMES Marine Cores Services. As we plan that transition, it will be most helpful to have feedback from our users, large and small. We therefore ask you to complete this short questionnaire and return it as soon as possible to one or both of  Ian Robinson (isr@noc.soton.ac.uk), or GHRSST-PP International Director, Craig Donlon (craig.donlon@metoffice.gov.uk).

Download this questionnaire in pdf or word format.

29 February 2007 L4 Galapagos/Coco Islands
A new high resolution multi-sensor product is now available over Galapagos and Coco Islands.
These 2km resolution maps are produced daily, merging observations from NOAA-17 and NOAA-18 AVHRR, ENVISAT/AATSR, AQUA/AMSRE and TRMM/TMI. They are being produced in the frame of an ESA project about biodiversity during a one month demonstration period (until 31st of March). The method used for the image reconstyruction and the data format are identical to the already existing L4 products over Mediterranean and North-West Europe areas.
The data can be accessed on ftp at :

ftp://ftp.ifremer.fr/ifremer/medspiration/data/l4uhrsstfnd/eurdac/gal/
30 January 2007 L2P AMSRE, TMI, AVHRR GAC and LAC Cessation of AMSRE, TMI, AVHRR GAC and LAC production
At the last Medspiration user consultation meeting (Villefranche-Sur-Mer, France, December 2006) discussions indicated that there was an unnecessary duplication of L2P processing activities within he GHRSSTPP system. In summary, Remote Sensing Systems (RSS) Inc. had recently begun to produce global operational feeds of L2P core swath and gridded format SST data from both AMSR-E and TRMM TMI, and NAVOCEANO had been producing AVHRR LAC and GAC data for a few month. Medspiration produce limited Atlantic area L2P feeds from gridded RSS AMSR-E and TMI fields and swath AVHRR LAC and GAC. 
Following some discussion, it was agreed that Medspiration should stop production of AMSR-E, TMI, AVHRR17 GAC and LAC L2P data feeds as these are no longer required by the user community due to the superior product provided from RSS and NAVOCEANO. The production of these products will thus be stopped on 27th of February.
Users are encouraged during this time frame to plug their system to the new repositeries for these datasets. They are all available at US-GDAC (ftp://podaac.jpl.nasa.gov/pub/GHRSST/data/L2P/) or mirrored on  European GDAC repositery (ftp://ftp.ifremer.fr/ifremer/medspiration/data/l2p/usdac).
* AMSRE and TMI L2P data are now in both swath and gridded format  but in L2P core format instead of the full-compliant L2P format of Medspiration. This means they lack the following variables : sea_ice_fraction, aerosol_optical_depth, DT_analysis, surface_solar_irradiance. The definition of values for proximity_confidence, rejection_flag, confidence_flag is also different (check netcdf file metadata for more information). They are two different products : one real-time (contains the _rt_ chain in filename) and one delayed mode available a few days after acquisition. Only the real-time product is mirrored on European GDAC repositery.
* AVHRR LAC and GAC products are now in swath format (the Medspiration files where 1D arrays with valid points only). They are available for both NOAA-17 and NOAA-18. The are also in L2P core format whereas Medspiration files where fully L2P-compliant. They lack the following variables : sea_ice_fraction, wind_speed, surface_solar_irradiance. The definition of values for rejection_flag is also different (check netcdf file metadata for more information).
09 January 2007 MDB The match-up database is now online!
Colocated in situ to satellite measurements can now be retrieved through the MDB extraction interface. Go to the following section.
09 January 2007 L4 Google vision now available!
The L4 products over Mediterranean Sea and North-West Europe can now be visualized through Google. Go to the following section.
21 September 2006 L2P AATSR The errors associated with each proximity confidence flag were changed as follow :

Proximity confidence bias standard deviation
2 (bad) 0.19 (formerly 0.33) 0.6
3 (suspect) 0.19 (formerly 0.33) 0.3
4 (acceptable) 0.19 0.3
5 (excellent) 0.19 0.3
6 (suspect, cold skin, upwelling, riverine input, etc...) 0.19 0.3

The justification for this request comes from the attached report from UK MetOffice who have performed sensitivity studies on the effect of the bias level in their OSTIA product.  (see last paragraph of section 3 of the report)  It is also endorsed by the AATSR validation scientist (Gary Corlett) whose name is also on the document.  Although there are some aspects of the paper that need to be clarified, the main conclusion is that the level of 0.19 bias assigned to good data (PCV levels 4 & 5) gives quite good agreement with in situ observations, but the value of 0.33 K presently used for levels 2 and 3 is biasing the response high whenever data with these PCV levels are used..  This was particularly noticeable up to June before we changed the criteria for determining the PCV levels and too many data were being assigned level 3 or 2.  Although the problem should be reduced since we changed those criteria, the paper shows that it is not wise to assign arbitrarily the value of 0.33 K bias to PCV 2 and 3.  The more sensible thing to do, until we can assign biases on the basis of observational evidence, is to use the same bias value for all PCVs. Read the report
9 August 2006 L2P AATSR The estimation of the proximity confidence flag was changed (threshold for distance to cloud updated) following new inputs from AATSR Validation Team of University of Leicester. The associated bias and standard deviation are unchanged but wil be updated later. The proximity are now estimated as follow  :

Distance to minimum climatology (in K)
> 1.5 3
(suspect)		 3
(suspect)		 5
(excellent)
1.5 to 0. 2
(bad)		 2
(bad)		 4
(acceptable)
0. to -1 1
(cloudy)		 2
(bad)		 4
(acceptable)
< -1. 1
(cloudy)		 1
(cloudy)		 6
(suspect, cold skin, upwelling, riverine input, etc...)
<1.
(close)		  1. to 3.
(near)		 >3.
(far)		 distance from nearest cloudy pixel (in km)
5 June 2006 L2P AATSR The estimation of the proximity confidence flag was changed (threshold for distance to cloud updated) following outcomes of Medspiration User Meeting. The proximity are now estimated as follow  :

Distance to minimum climatology (in K)
> 1.5 3
(suspect)		 3
(suspect)		 5
(excellent)
1.5 to 0. 2
(bad)		 2
(bad)		 4
(acceptable)
0. to -1 1
(cloudy)		 2
(bad)		 4
(acceptable)
< -1. 1
(cloudy)		 1
(cloudy)		 6
(suspect, cold skin, upwelling, riverine input, etc...)
<2.
(close)		  2. to 5.
(near)		 >5.
(far)		 distance from nearest cloudy pixel (in km)

The associated bias and standard deviation are :
proximity confidence bias standard deviation
2 (bad) 0.33 0.6
3 (suspect) 0.33 0.3
4 (acceptable) 0.19 0.3
5 (excellent) 0.19 0.3
6 suspect, cold skin, upwelling, riverine input, etc...) 0.19 0.3


18 January 2006 L4 UHR The configuration of the L4 Ultra-high resolution product over Mediterranean sea is changed :
  • area now including MFSTEP grid ([30.N, 46.5N], [18.5W, 36.5E])
  • analysis centred on 0:00 UTC of current day instead of 12:UTC of previous day
  • use night data only
  • use simple mean for data centering instead of Bretherton mean
  • no more wind threshold in the input data selection
  • generation time : 18:00 UTC (instead of 7:00) since we take into account the data of the current morning (until 12:00)
6 December 2005 L2P AATSR The AATSR is now processed over global ocean (instead of Atlantic), following request from ESA.
25 November 2005 L2P NAR16 Switch from NOAA-16 to NOAA-18 for NAR product : The production of NAR16 L2P was stopped and replaced by NAR18 product, following decision by the SAF O&SI to move from NOAA-16 (questionable data quality) to NOAA-18.
3 August 2005 All L2P Faugere Climatology switched to USGODAE FNMOC (GHRSST Analysis) : The source for the DT_analysis field computation in the L2P consolidation was changed to USGODAE/FNMOC analysis instead of Faugere Climatology, as a recommendation by GHRSST Project Office. If the source for USGODAE analysis is unavailable for more than 3 days, the system is switched back to Faugere climatology.
21 June 2005 All L2P Switch from NISE (NSIDC) to SAF O&SI sea-ice product : The source for sea ice concentration ancillary field in the L2P consolidation was changed to SAF O&SI product instead of NSIDC data. The new SAF (Met.No) product is global, widely assessed and produced operationally in a sustained manner. The NISE product is still collected in case of unavailability of SAF O&SI product and the system can be switched back to NISE anytime.
7 June 2005 L2P NAR16, NAR17, NAR18 Extension of NAR product coverage : As agreed at the first Medspiration User meeting, the coverage of the NAR products was extended to the full coverage of native SAF grid.
7 June 2005 All L2P Switch from NOAA aerosols (AVHRR NESDIS) to NAAPS aerosols : The source for aerosol optical depth ancillary field in the L2P consolidation was changed to NAAPS product instead of AVHRR NESDIS product, as a recommendation by the GHRSST Project Office.

Satellites